Studying accelerated cardiovascular ageing in Russian adults through a novel deep-learning ECG biomarker

Background: A non-invasive, easy-to-access marker of accelerated cardiac ageing would provide novel insights into the mechanisms and aetiology cardiovascular disease (CVD) as well contribute to risk stratification those who have not had a heart or circulatory event. Our hypothesis is that differences between an ECG-predicted chronologic age participants (?age) reflect decelerated ageing Methods: convolutional neural network model trained on over 700,000 ECGs from Mayo Clinic in U.S.A was used predict 4,542 Know Your Heart study conducted two cities Russia (2015-2018). Thereafter, ?age linear regression models assess associations with known CVD factors markers abnormalities. Results: The biomarker (mean: +5.32 years) strongly positively associated established for CVD: blood pressure, body mass index (BMI), total cholesterol smoking. Additionally, strong independent positive structural abnormalities: N-terminal pro b-type natriuretic peptide (NT-proBNP), high sensitivity troponin T (hs-cTnT) pulse wave velocity, valid vascular ageing. Conclusion: difference ECG-age obtained contains information about level exposure individual damage way consistent it being (vascular) ageing. Further research needed explore whether these are seen populations different risks events, better understand underlying involved.